elvindp
elvindp
in 'mmdet/models/anchor_heads/solov2_head.py', function 'solov2_target_single' generate ground truth label, but I find that, when initial cate_label, the value is 0: cate_label = torch.zeros([num_grid, num_grid], dtype=torch.int64, device=device) and then update cate_label, if...
Hi, in the implementation, a mask with input size is used, which is confused to me. In NLP transformer, mask is used to avoid influence of different input length zero...
Hi, thanks for your great work! I have some questions about the code: Becuase without the score, if the target is out of FOV (i.e. out of image), the lightning...
in 'mmdet/models/anchor_heads/solov2_head.py', function 'solov2_target_single' generate ground truth label, but I find that, when initial cate_label, the value is 0: cate_label = torch.zeros([num_grid, num_grid], dtype=torch.int64, device=device) and then update cate_label, if...
在'solov2_head.py'里，'solov2_target_single'函数会生成ground truth label， cate_label初始值是0： cate_label = torch.zeros([num_grid, num_grid], dtype=torch.int64, device=device) 然后在下边根据中心点的位置，更新cate_label: cate_label[top:(down+1), left:(right+1)] = gt_label 但class是从0-80，第0类是人，物体很多时候并不是布满整幅图的，这样会cate label里就有大量的0类 有人可以帮忙解释一下不？谢谢
在论文里，提到用kalman预测了tracks的新位置，14-16行。 但在代码里，并没有用到预测的新位置。 `# Predict the current location with KF` ` STrack.multi_predict(strack_pool)` ` dists = matching.iou_distance(strack_pool, detections)` multi_predict用kalman预测了新的mean，covariance，但在iou_distance的计算里面，用的是旧的tlbr。
Hi, when l use the example python api to inference a video with 15mins, it consume large GPU memory and report torch.cuda.OutOfMemoryError. I find the script "load_video" extract 1 frame...